SP476
INDIVIDUALIZED APPROACH TO THE PRESCRIPTION OF BICARBONATE (HCO_3-) HEMODIALYSIS (HD): 12-MONTH PROSPECTIVE STUDY Free

INTRODUCTION AND AIMS: Metabolic acidosis is common in end-stage renal disease and HD patients. There is a variety of physiological factors affecting acid-base balance in those patients. The KDOQI guidelines therapeutic goal is to maintain pre-dialysis HCO_3-≥22mmol/L. The aim of the study was to evaluate an individualized HCO_3- HD prescription as a preventing factor of metabolic changes in a hospital HD facility and its influence on other treatment components.

METHODS: 12-month prospective study of patients on online high-flux hemodiafiltration and serum HCO_3- >30mmol/L. Every 3 months (5 time points: 1, 2, 3, 4 and 5) HCO_3-, Calcium (Ca²⁺) , Phosphorus (P⁺) and intact Parathyroid hormone (iPTH) blood levels were analyzed and hemodialysis HCO_3- was changed using the following rules: HCO_3- > 30 mmol/L reduce 4 mmol/L HCO_3- prescription; HCO_3- ≥ 25 mmol/L reduce 2 mmol/L HCO_3- prescription; 20 mmol/L <HCO_3- < 25mmol/L no change; HCO_3- ≤ 20mmol/L increase 2 mmol/L HCO_3-prescription; HCO_3- < 18mmol/L increase 4 mmol/L HCO_3- prescription. Data collected comprised demographic information, renal disease etiology, hydration status by bioimpedance, comorbidities, HD treatment information and lab results. Statistical analysis was made using SPSS version 23 for Mac OS X.

RESULTS: 34 patients: 61.8% males (n=21), 44.1% (n=15) diabetic and 64.7% (n=22) hypertensives. At baseline, average pH was 7.37±0.06 and HCO_3- 26.19 ±2.03mmol/L. At time point 5, pH was 7.36±0.07 and HCO_3- 23.87±1.93mmol/L.A repeated measures ANOVA with a Huyn-Feldt correction determined that HCO_3- prescription differed with statistical significance during time (p=0.001) and that DM (p=0.019), Hypertension (p=0.004) and renal disease etiology (p=0.004) influenced the prescription. Post hoc tests confirmed those assumptions. As expected, serum HCO_3- was influenced by time (p=0.001).Baseline serum HCO_3- was not associated with demographics. A Spearman’s rank-order correlation was run and determined moderate negative correlations between HCO_3- and iPTH (r_s=-0.49, p=0.004) and nPCR (r_s=-0.53, p=0.001). The Pearson’s correlation run at the same time found a moderate negative correlation between HCO_3- and P⁺ (r=-0.49, p=0.003) and a strong negative correlation with pre-HD urea (r=-0.59, p=0.001).Using T-student test we observed that, at baseline, diabetic had lower HCO_3- (time point 1: t=-2.4, p=0.022; time point 2: t=-2.2, p=0.034), but with the changes in the HCO_3- prescription those differences disappear in time points 3, 4 and 5. Using a Pearson’s correlation, we determined a moderate negative correlation between overhydration (OH) and HCO_3- at time points 2 to 3 (r=-0.4, p=0.029) and 4 to 5 (r=-0.4, p=0.023). That was the only association with bioimpedance measurements that we found.

CONCLUSIONS: Pre-dialysis serum HCO_3- is highly variable. The HCO_3- prescription was influenced by comorbidities like DM and Hypertension and renal disease etiology, however the same factors did not seem to be associated with serum HCO_3-. Laboratory values like iPTH, Ca²⁺, P⁺, Albumin, Protein C Reactive or Urea were also not associated with serum HCO_3-. From the bioimpedance evaluations, only the OH was associated with HCO_3-. We hypothesized that dialysis HD parameters would be important determinants of bicarbonate variability, but our findings did not support this.